When Will Britain Learn How to Manage Floods?
(15 December 2015)
Following the recent floods in Cumbria, Dr Jeff Warburton and colleagues from universities of Liverpool, Southampton and King's College London, consider the best way to manage future floods.
Yet again, distressing images of flood damage and destruction in Cumbria, northern England are prompting calls for further investment in UK flood defences and fearful talk of climate change.
There is a particularly worried tone to the commentary, because this flooding occurred after installation of new hard flood defences (completed in June 2010) that simply could not cope with the flow and - despite millions spent on glass flood walls and hydraulic modelling of the options - these defences failed. Or at least this is what we are hearing in the press.
This verdict is too severe, flood defences have worked in places and have bought people - and the emergency services - more time and reduced flood risk.
But for those with homes covered in wet, polluted mud, these arguments sound hollow. The simple fact is that with record-breaking rainfall (341mm in 24 hours at Honister Pass) on a landscape which has been managed for centuries to be efficient at encouraging rainwater to flow into rivers, you would expect flooding - and the more extreme the rainfall, the more extreme the flooding. A key question being asked by all is: "What do we mean by extreme flooding?"
Already we have heard repeatedly the common misconception "this is a flood that should only happen every 100 years". This is incorrect. The science of flood extremes attempts to use the woefully short records of river flow (which has been around for about 30 years) and rainfall (which has been recorded for about 200 years) to estimate a probability: the chance of a flood of a given size happening.
A 100-year flood is actually better described as a flood that has a 1% chance of occurring within any given year. It does not mean that we expect one every 100 years - sadly as the people of Cumbria have found, a flood with a 1% chance of occurring can occur more than once in 100 years and quite possibly within the same year. In fact the chances of another 100-year event occurring in 2016 remains 1%.
We've been here before
After the 2009 Cumbrian floods, the Natural Environment Research Council funded a study to look at using lake sediments to track any changes in flood frequency and magnitude. The lake sediment record from our lakes can provide a long perspective on flood magnitude and frequency spanning thousands of years.
The mud covering the flooded homes and roads have been washed down rivers from the mountains - and in Cumbria they end up in the lakes. Each flood can leave a distinct sediment layer. Research by scientists from the universities of Southampton, Liverpool and Durham suggests that the floods of 2005 and 2009 in Cumbria were the largest for 600 years (the results of this analysis are yet to be published) and that these sorts of events had a probability of occurring 0.001% or one in every 1,000 years (substantially rarer than the one in 200-year value arrived at using the 30 years of measured river flows).
What is worrying is that the lake sediment flood record shows that two-thirds of the very largest floods experienced in Cumbria have happened in the past 15 years, a period characterised by warmer northern hemisphere temperatures (an index for moisture and energy in the atmosphere) and positive North Atlantic Oscillation Index - a measure of the direction of storm tracks over the UK (including Storm Desmond).
Such conditions last occurred 800 years ago during the medieval climatic anomaly, a time when the Cumbria landscape was quite different from today with less intensive agriculture and more extensive woodlands.
In the Howgill Fells a lack of trees and steep slopes have created conditions for landslides - sediment from which is then fed into the river channel creating changes in the river. A wooded valley would make the slopes less susceptible to landslides and would reduce the movement of sediments and water downstream. As it is, the huge rainfall that came with Storm Desmond has resulted in landslides around the valleys of Borrowdale, Thirlmere, Ullswater and the Howgill Fells.
In 2009, the extreme rainfall caused fewer landslides than the rainstorms of 2005 - this is largely due to the fact that many slopes had already failed in the previous flooding episodes so that in 2009 far fewer slopes were vulnerable to failure. What we need to take away from this is that the state of the landscape on which extreme rains fall will determine what happens. This underlines a need for landscape management as well as engineered flood defences.
Needed: more trees
Cumbria 2015 has brought this home with a vengeance. We need to seriously start acting to manage both landscapes, rivers and streams to release the water more slowly. There is growing evidence that we need strategic planting of trees in our uplands and to restore more natural rivers that connect better with floodplains. These approaches have dual benefits - reducing the transport of sediment down into critical flood channels and reducing peak river levels by slowing flows.
The main difficulty here will be the complex jigsaw of organisations and individuals involved in managing land and water. A key challenge will be to find effective ways to encourage planners, farmers and town councils to work together to adapt our catchments to farm for food production, but also manage water and sediment yield to rivers. Flooding is a connecting process - the soil and water in the homes in Carlisle came from the hills and valleys of the Lake District.
Perhaps the sequence of floods in Cumbria will help those in power to change their view of large floods - yes they are devastating, but maybe they can also act as a catalyst for change that results in better landscapes for our environment and more connected approaches to flood risk management - not just bigger flood defences.
First published in The Conversation.